14 Supported AI Providers

Connect your own API keys (BYOK — Bring Your Own Key). Your keys stay AES-256-GCM encrypted on our servers. Supporters send you tokens, and your calls are automatically routed through those token balances.

Claude

Anthropic

Anthropic's most capable models. Best for complex reasoning, coding, and analysis.

claude-opus-4-7
claude-sonnet-4-6
claude-haiku-4-5-20251001

GPT

OpenAI

OpenAI's GPT models. Industry standard for chat and code completion.

gpt-4o
gpt-4o-mini
gpt-4.1
gpt-4.1-mini
o4-mini

Gemini

Google

Google's multimodal models with long context windows.

gemini-2.5-pro
gemini-2.5-flash
gemini-2.0-flash

OpenRouter

OpenRouter

Access 200+ models from every major provider through a single API.

All 200+ models via routing

Groq

Groq

Blazing fast inference with Groq's LPU chips. Best for low-latency tasks.

llama-3.3-70b-versatile
llama-3.1-8b-instant
mixtral-8x7b-32768

Grok

xAI

xAI's Grok models with real-time web access and advanced reasoning.

grok-3
grok-3-mini
grok-2-1212

Mistral

Mistral AI

European frontier AI with strong coding and multilingual capabilities.

mistral-large-latest
mistral-small-latest
codestral-latest

DeepSeek

DeepSeek

State-of-the-art open models with exceptional reasoning at low cost.

deepseek-chat
deepseek-reasoner

Cohere

Cohere

Powerful enterprise-focused models optimized for RAG and search.

command-r-plus
command-r
command-a-03-2025

Perplexity

Perplexity AI

AI models with real-time web search and citations built in.

llama-3.1-sonar-large-128k-online
sonar-pro

Together AI

Together AI

Fast open-source model inference with 100+ models available.

meta-llama/Llama-3.3-70B-Instruct-Turbo
Qwen/Qwen2.5-72B-Instruct-Turbo

Fireworks

Fireworks AI

Production-grade inference for open models with function calling.

accounts/fireworks/models/llama-v3p1-70b-instruct
firefunction-v2

Cerebras

Cerebras

World's fastest AI inference on Cerebras Wafer-Scale Engine chips.

llama3.3-70b
llama3.1-8b

AI21 Labs

AI21

Jamba models with hybrid SSM-Transformer architecture for long context.

jamba-1.5-large
jamba-1.5-mini

How the API proxy works

1. Connect your API key in the dashboard (stored AES-256-GCM encrypted — never exposed).

2. Use your GMT key (gmt_xxx) in any OpenAI-compatible tool.

3. Requests hit https://givemesometokens.dev/api/v1/chat/completions.

4. The proxy decrypts your stored key, forwards the request to the correct provider, and deducts tokens from your wallet.

5. Supporters' token donations become your API budget — no API key is ever shared.